- Email: [email protected]
Snail expression is an independent predictor of tumor recurrence in superficial bladder cancers
Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
Original article
Snail expression is an independent predictor of tumor recurrence in superficial bladder cancers夡 Franck Bruyere, M.D.a,1, Benjamin Namdarian, M.B.B.S.b,1, Niall M. Corcoran, Ph.D., M.D.b, John Pedersen, M.D.c, Jeremy Ockrim, M.D.d, Bryan B. Voelzke, M.D.e, Uttam Mete, M.D.f, Anthony J. Costello, M.D.b, Christopher M. Hovens, Ph.D.b,* a Department of Urology, CHU Bretonneau, Tours, France Departments of Urology and Surgery, Royal Melbourne Hospital, University of Melbourne, Parkville, Australia c TissuPath Pty Ltd., Hawthorn, Australia d Departments of Urology, Imperial College School of Medicine, Hammersmith Hospitals NHS Trust, London, United Kingdom e University of California, San Francisco, CA 94143, USA f Postgraduate Institute of Medical Education and Research, Chandigarh, India b
Received 21 August 2008; received in revised form 6 November 2008; accepted 9 November 2008
Abstract Background: Epithelial-mesenchymal transition (EMT) is known to play an important role in the development of tumor invasion and progression in tumors of epithelial origin. Objectives: Our aim was to investigate the role of Snail transcription repressor family members in human bladder pathogenesis. Material and methods: We evaluated levels of Snail and Slug in 87 patients who received transurethral resection of a transitional cell carcinoma at our institution during the period from June 1999 until November 2003. Immunohistochemistry was performed on tissue microarrays, and expression correlated with pathological variables and clinical outcomes. Degree and intensity of Snail and Slug staining was quantified by immunohistochemistry. Results: There was no apparent enrichment in strong vs. weak staining for either Snail (43.7% vs. 56.3%) or Slug (46% vs. 54%) in the superficial bladder tumors. Univariate analysis determined that tumor focality and Snail expression were significantly associated with tumor recurrence (P ⬍ 0.05). Only for tumor focality did such a relationship exist when assessing tumor progression. Multivariate analysis using the Cox’s proportional hazards model revealed similar results to that of the univariate analysis. Snail expression (P ⫽ 0.038) and tumor focality (P ⫽ 0.011) were independent and significant prognostic factors for tumor recurrence in all patients. However, only tumor focality was an independent predictor of tumor progression (P ⫽ 0.034). Conclusions: High expression of Snail in superficial bladder tumors is a strong predictor of tumor recurrence enhancing risk stratification and prognostication. © 2010 Elsevier Inc. All rights reserved. Keywords: Snail; Slug; Immunohistochemistry; Recurrence; Carcinoma; Transitional cell
1. Introduction
夡 This work was supported by funds from the Melbourne Urology Trust, by a generous gift from Roger Riordan from the Cybec Trust. N.M.C. was supported by a Benjamin Rank Surgical Fellowship from the Royal Melbourne Hospital, and a NHMRC Medical Postgraduate Scholarship. B.N. was supported by the 2007 Raelene Boyle Scholarship from the Royal Australasian College of Surgeons. * Corresponding author. Tel.: ⫹613-9342-7703/4; fax: ⫹613-9347-6488. E-mail address: [email protected] (C.M. Hovens). 1 These authors contributed equally to this work.
1078-1439/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2008.11.005
Cases of bladder cancer contribute significantly to the overall tumor burden with more than 60,000 new cases diagnosed in the United States in 2006, with approximately 13,000 deaths from the disease [1]. The most common bladder cancer histological subtype is that of transitional cell carcinoma, with some 70% to 80% of these cases initially presenting with superficial tumors that are confined to the epithelium or subepithelial connective tissue. Transurethral surgical resection of these tumors remains the treat-
592
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
ment of choice, however 31% up to 78% of these cases will recur [2]. Of those that recur, up to 45% progress to invade the underlying detrusor muscle, which is associated with a significantly poorer outcome [3,4]. Current pathological classification schema cannot adequately predict tumor behavior nor permit patient selection for early aggressive treatment or long-term surveillance as appropriate. Molecular profiling of key tumor signaling and regulatory proteins in bladder cancer may lead to improved tumor classification and clinical outcomes. Transition from an epithelial to mesenchymal state (EMT) is a necessary event prior to tumor cells of epithelial origin migrating and metastasizing to secondary tissues [5–7]. Loss of E-cadherin mediated cell adhesion is a common event in epithelial cancers and has been correlated with dedifferentiation, a higher risk of local invasion and metastatic disease, and poor prognosis in a variety of tumors of epithelial origin [8 –12]. Recent attention has focused on transcription factors of the Snail superfamily, Snail [13] and SLUG [14], which can suppress transcription from the E-cadherin promoter. Expression of the Snail superfamily transcription factors has been found to negatively correlate with E-cadherin expression in various epithelial cancers [15–18]. The aims of this study were to analyze the expression levels of the Slug and Snail transcription repressor proteins family members in superficial bladder tumor specimens and to correlate these levels with clinical and pathological features and oncological outcomes.
2. Materials and methods 2.1. Protocol In our study we assessed 87 patients (70 men and 17 women) with a mean age of 71 years during the period from June 1999 until November 2003, who received transurethral resection of a transitional cell carcinoma at our institution. Approval for the study was granted by the Research Directorate of Melbourne Health, Australia. Samples of the tumors that had been fixed in 10% formalin and paraffin embedded shortly after the procedure were obtained from the archives of the Department of Pathology. Sections of the specimens were stained with hematoxylin-eosin (HE) and analyzed by two independent practitioners, classifying each according to TNM and WHO criteria. 2.2. Tissue microarrays (TMA) generation From the HE-stained sections tumor regions were identified and 2 to 3 tissue cylinders of diameter 1.5 mm were taken and layered into a recipient block using a tissue chip microarrayer. The recipient block was subsequently sec-
tioned onto Histobond (Leica, Marienfeld, Germany) sialinized glass slides. Graded ethanol solutions were used to deparaffinize and rehydrate the 5 m paraffin-embedded specimen sections while 0.3% hydrogen peroxide was used to block endogenous peroxidase activity. Antigen retrieval was performed by heating the slide in a microwave oven in EDTA buffer (10 mM; pH 8, 7 minutes). Sections were cooled at room temperature for 30 minutes before incubated with primary antibodies at 4oC. The following antibodies were used: SNAI 1 (H-130):sc-28199 (1:400); Santa Cruz Biotechnology (Santa Cruz, CA), SLUG (D-19):sc-10437 (1:100); Santa Cruz Biotechnology. Following primary antibody application, the sections were incubated for 30 minutes with peroxidase labeled polymer (DAKO Cytomation Envision⫹ dual link system; Carpinteria, CA). Sections were subsequently washed thoroughly in phosphate-buffered saline, and 3,3=-diamino-benzidine (DAKO Cytomation Liquid DAB⫹) was used to develop the peroxidase signal. In assessing the slides, two independent observers blindly examined the staining in TMAs, and a mean score was calculated. The positive reaction was scored into 4 categories for both the intensity of staining and the percentage of positive cells. Grades according to the intensity of the staining included: 0, 1⫹, 2⫹, and 3⫹ while the percentages of positive cells were separated into: 0 (0%), 1 (1% to 33%), 2 (34 to 66%), and 3 (67 to 100%). In order to maintain consistency between observers, when the rare occurrence of a discrepancy arose, a third observer gave the final score. In the final statistical analysis, the staining results were classified into 2 groups depending on the intensity and the percentage of positive cells. Weak staining (intensity 0 or 1, or more but percent category 0 or 1); and strong staining (intensity 2 or 3 and percent category 2 or 3). 2.3. Statistical analysis In performing statistical analyses with SPSS version 9 (Chicago, IL), age, sex, stage, grade, multifocality, tumor weight, and antibody overexpression were the tumor and patient variables analyzed. Chi Square test or Fisher’s exact test were used to demonstrate any intercorrelations between stage, grade, multifocality, tumor focality, and antibody over expression. Time to recurrence and progression were analyzed using the Kaplan-Meier method. Correlations between tumor recurrence or progression, and antibody overexpression, as well as pathological variables such as grade and focality were identified using the log rank test. The date of the last cystoscopy was used to define the periods of assessment for patients without recurrence, progression, or cancer death. The Cox proportional hazards model was used for multivariate analysis of parameters predictive of tumor recurrence and to estimate relative risks. P values reported in the article are all 2-sided, significant at P ⬍ 0.05.
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
593
Table 1 Clinical and histological data of 87 patients with superficial bladder cancer
Table 3 Tumor recurrence data for Slug and Snail proteins for 87 superficial bladder tumor specimens
Number Sex (M:F) ratio Mean age (y) SD Recurrence (%) Progression (%) Median follow-up Grade (%) PUNLMP Low grade High grade Uni/multifocal Median Tumor weight (range) g
Variable
87 70/17 71 (30–96) 50 (57.5%), mean 14.8 mo 11 (12.6%), mean 16.3 mo 33.9 mo (0.5–66) 12 (13.8%) 61 (70.1%) 14 (16.1%) 61.8 %/38.2% (11 missing) 3 (1–55)
3. Results 3.1. Clinicopathologic summary
Grade PUNLMP Low High Focality Unifocal Multifocal Weight ⬍Median ⬎Median Slug Poor Strong Snail Poor Strong
Number recurred
Mean time to recurrence (SE)
9/12 33/61 8/14
28.3 (16.7–39.8) 29.4 (23.3–35.6) 29.5 (15.9–43.1)
23/47 20/29
33.6 (27.0–40.2) 19.9 (11.5–28.3)
28/46 22/41
29.4 (22.6–36.2) 27.7 (21.0–34.4)
26/47 24/40
27.7 (21.4–34.1) 29.5 (22.2–36.8)
24/49 26/38
33.6 (27.0–40.2) 22.6 (15.7–29.6)
Lag rank
P value
0.04
0.9795
7.34
0.0068
0
0.9457
0
0.994
4.85
0.0275
Cox proportional hazards model
A total of 87 consecutive cases of superficial urothelial transitional cell carcinoma of the bladder were selected in this study (Table 1). All tumors were classified as superficial pTa tumors according to the TNM criteria, and subgrouped into grades by the World Health Organization (WHO) criteria with papillary urothelial neoplasms of low malignant potential (PUNLMP) comprising 12 patients, low grade consisting of 61 patients, and high grade consisting of 14 patients (Table 1). Interestingly, we were unable to find a correlation on univariate or multivariate analyses between tumor grade (Table 2) and recurrence (Table 3) or progression (Table 4), which we believe reflects disease presentation in our cohort, the current WHO grading system placing a larger proportion of cases into the low grade category, hence limiting the power of any statistical comparison to PUNLMP or high grade tumors. Further clinical and histologic data are summarized in Table 1. 3.2. Tissue immunohistochemistry analysis We examined the expression of both Snail and Slug in the 87 tumor samples (Fig. 1). There was no apparent enrichment in strong vs. weak staining for either Snail (43.7% vs. 56.3%) or Slug (46% vs. 54%) in the superficial bladder tumors (Fig. 2). We then correlated protein expres-
Table 2 Pearson 2 odds ratio correlation between Slug and Snail expression and standard bladder tumor prognostic indicators
Slug Snail
Weight
Focality
Grade
1.2 (0.5–2.9) P ⫽ 0.62 1.48 (0.63–3.47) P ⫽ 0.365
2.3 (0.9–6.1) P ⫽ 0.087 0.91 (0.36–2.32) P ⫽ 0.846
NA P ⫽ 0.936 NA P ⫽ 0.54
Variable
HR (95% CI)
P value
Grade Weight Focality Slug Snail
1.06 0.93 2.25 1.07 1.87
0.83 0.81 0.011 0.83 0.038
(0.65–1.73) (0.53–1.65) (1.20–4.21) (0.58–1.98) (1.04–3.38)
sion with established histological predictors of bladder tumor aggression, including grade, tumor bulk, and focality. We established on Pearson 2 odds ratio correlation that neither Snail nor Slug expression levels correlated significantly with any of these histologic predictive parameters (Table 2). We then proceeded to examine on univariate and multivariate analysis whether protein expression of Snail or Slug correlated with patient prognosis in regards to tumor recurrence and progression in all patients (n ⫽ 87). The univariate analysis using the log rank test determined that tumor focality (P ⫽ 0.0068) and Snail expression (P ⫽ 0.0275) were the only parameters significantly associated with tumor recurrence (Table 3). The predictive ability of Snail is demonstrated in Fig. 3, showing the Kaplan Meier survival curve. Multivariate analysis using Cox’s proportional hazards model revealed that only Snail expression (P ⫽ 0.038) and tumor focality (P ⫽ 0.011) were independent and significant prognostic factors for tumor recurrence in all patients (Table 3). We then determined if expression of these transcriptional repressor proteins, influenced tumor progression. We found that tumor focality (P ⫽ 0.0237) but neither Snail nor Slug levels significantly predicted for tumor progression on univariate analysis (Table 4). Upon multivariate analysis, again only tumor focality was an independent predictor of tumor progression (P ⫽ 0.034, Table 4).
594
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
Table 4 Tumor progression data for Slug and Snail proteins for 87 superficial bladder tumor specimens Variable Grade PUNLMP Low High Focality Unifocal Multifocal Weight ⬍Median ⬎Median Slug Poor Strong Snail Poor Strong
Number progressed
Mean time to progression (SE)
3/12 5/61 3/14
49.1 (38.5–59.6) 60.9 (56.7–65.2) 50.4 (38.9–61.9)
3/47 7/29
58.5 (54.6–62.3) 51.9 (42.2–60.6)
5/46 6/41
60.1 (55.2–64.9) 49.1 (43.4–54.9)
6/47 5/40
58 (52.1–63.9) 53.8 (48.8–58.9)
5/49 6/38
53.2 (49.2–57.1) 56.3 (49.2–63.4)
Log rank
P value
2.9
0.234
5.12
0.0237
0.71
0.4
0.04
0.843
0.71
0.399
Cox proportional hazards model HR (95% CI)
P value
Grade Weight Focality Slug Snail
0.75 1.64 4.56 1.08 1.55
0.63 0.5 0.034 0.9 0.5
(0.23–2.43) (0.49–5.47) (1.12–18.57) (0.31–3.82) (0.44–5.29)
4. Discussion Currently, decisions regarding treatment regimens for urothelial carcinoma are based solely on the TNM staging schema. This classification criteria is well suited to describe tumor morphology, however, this information cannot be extrapolated to provide a full assessment of the risk for individual tumor recurrence, progression, or response to a given course of treatment. Panels of molecular biomarkers, which have the potential to be objectively quantified and routinely incorporated into pathological analyses, might overcome current morphological assessments and permit eventual stratification of clinically relevant risk profiles of pathologically similar bladder tumors. We determined the expression profiles of Snail and Slug and their relationship to the clinicopathological features of the superficial bladder tumors. We found that elevated levels of Snail expression were significantly associated with bladder tumor recurrence, which is suggestive that this marker may be an indicator of poor prognosis in superficial bladder cancer. Surprisingly, we found no such correlation with the Snail related transcriptional repressor, Slug. The finding that Snail expression is associated with tumor recurrence conforms to the hypothesis that transcriptional regulators of EMT are important for advanced tumor features such as invasion and metastasis. Interestingly, in their normal physiologic context, Snail genes appear to control cell movement processes that do not require a full EMT
Fig. 1. Representative images of immunohistochemical staining of the indicated antibodies on superficial bladder cancer microarrays. Magnification at 100⫻ to 400⫻. (Color version of figure is available online.)
such as mesoderm formation by regulating cell adhesion and migration [19]. Snail expression is positively correlated with poor outcome in metastatic breast cancer patients, with Snail RT-PCR positive breast carcinoma effusions associAntibody Slug Snail
Strong Staining 46 43.7
Weak Staining 54 56.3
100% 90% 80% 70% Percentage Staining
Variable
60% Weak
50%
Strong
40% 30% 20% 10% 0% Slug
Snail
Fig. 2. Expression of Snail and Slug antibodies as assessed by immunohistochemical staining of the superficial bladder tumor cohort. (Color version of figure is available online.)
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
595
[16]. Overall, these findings suggest that different members of the Snail transcription repressor family are likely to have highly specific roles in regulating EMT in different tumor contexts. A knowledge of these roles will need to be obtained before specific therapies, targeted to individual members of this transcriptional repressor family, can be effectively applied to treat different epithelial carcinomas.
5. Conclusions To date, no single biomarker has consistently enabled risk stratification and prognostication of superficial bladder tumors. When analyzing Snail and Slug expression levels, we have found that high expression of Snail in superficial bladder tumors significantly predicts tumor recurrence in these patients.
References
Fig. 3. Kaplan-Meyer curves of tumor recurrence analyzing Slug and Snail expression levels in superficial bladder tumors.
ated with extremely poor survival [20]. In addition, a recent analysis of 48 primary ovarian tumors and their corresponding metastases for expression of E-cadherin and Snail by immunohistochemistry found that Snail expression in metastases (P ⫽ 0.047) but not the primary tumor was significantly associated with reduced overall survival of the patients [17]. However, neither Snail nor Slug expression were positively correlated with metastatic effusions, disease-free survival, or overall survival in an ovarian cancer cohort [20]. In this tumor setting, it was the Snail transcription factor family member, Smad interacting protein one (sip1), that was associated with higher tumor stage, metastatic effusions and poorer overall survival [20]. In contrast to our findings with superficial bladder cancers, Slug up-regulation has been found to correlate with deeper invasion, lymph node metastases and tumor stage in esophageal squamous cell carcinoma [21], and with distant metastases and advanced tumor stage in gastric carcinoma
[1] Jemal A, Siegel R, Ward E, et al. Cancer statistics 2006. CA Cancer J Clin 2006;56:106 –30. [2] Babjuk M, Oosterlinck W, Sylvester R, et al. EAU Guidelines on Non-muscle-invasive urothelial carcinoma of the bladder. Eur Urol 2008;54:303–14. [3] Holmang S, Hedelin H, Anderstrom C, et al. The relationship among multiple recurrences, progression and prognosis of patients with stages Ta and T1 transitional cell cancer of the bladder followed for at least 20 years. J Urol 1995;153:1823– 6, Discussion 1826 –7. [4] Heney NM, Ahmed S, Flanagan MJ, et al. Superficial bladder cancer: Progression and recurrence. J Urol 1983;130:1083– 6. [5] Bagnato A, Rosano L. Epithelial-mesenchymal transition in ovarian cancer progression: A crucial role for the endothelin axis. Cells Tissues Organs 2007;185:85–94. [6] Moustakas A, Heldin CH. Signaling networks guiding epithelialmesenchymal transitions during embryogenesis and cancer progression. Cancer Sci 2007;98:1512–20. [7] Hugo H, Ackland ML, Blick T, et al. Epithelial-mesenchymal and mesenchymal-epithelial transitions in carcinoma progression. J Cell Physiol 2007;213:374 – 83. [8] Moll R, Mitze M, Frixen UH, et al. Differential loss of E-cadherin expression in infiltrating ductal and lobular breast carcinomas. Am J Pathol 1993;143:1731– 42. [9] Fujimoto J, Ichigo S, Hirose R, et al. Expression of E-cadherin and ␣and -catenin mRNAs in uterine cervical cancers. Tumor Biol 1997; 18:206 –12. [10] Jawhari A, Jordan S, Poole S, et al. Abnormal immunoreactivity of the E-cadherin-catenin complex in gastric carcinoma: Relationship with patient survival. Gastroenterology 1997;112:46 –54. [11] Jeffers MD, Paxton J, Bolger B, et al. E-cadherin and integrin cell adhesion molecule expression in invasive and in situ carcinoma of the cervix. Gynecol Oncol 1997;64:481– 6. [12] Huiping C, Kristjansdottir S, Jonasson JG, et al. Alterations of E-cadherin and -catenin in gastric cancer. BMC Cancer 2001;1:16. [13] Cano A, Perez-Moreno MA, Rodrigo I, et al. The transcription factor Snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression. Nat Cell Biol 2000;2:76 – 83. [14] Nieto MA. The Snail superfamily of zinc-finger transcription factors. Nat Rev Mol Cell Biol 2002;3:155– 66. [15] Rosivatz E, Becker I, Specht K, et al. Differential expression of the epithelial-mesenchymal transition regulators Snail, SIP1, and twist in gastric cancer. Am J Pathol 2002;161:1881–91.
596
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
[16] Castro Alves C, Rosivatz E, Schott C, et al. Slug is overexpressed in gastric carcinomas and may act synergistically with SIP1 and Snail in the down-regulation of E-cadherin. J Pathol 2007;211: 507–15. [17] Blechschmidt K, Sassen S, Schmalfeldt B, et al. The E-cadherin repressor Snail is associated with lower overall survival of ovarian cancer patients. Br J Cancer 2008;98:489 –95. [18] Becker KF, Rosivatz E, Blechschmidt K, et al. Analysis of the E-cadherin repressor Snail in primary human cancers. Cells Tissues Organs 2007;185:204 –12.
[19] Barrallo-Gimeno A, Nieto MA. The Snail genes as inducers of cell movement and survival: Implications in development and cancer. Development 2005;132:3151– 61. [20] Elloul S, Elstrand MB, Nesland JM, et al. Snail, Slug, and Smadinteracting protein 1 as novel parameters of disease aggressiveness in metastatic ovarian and breast carcinoma. Cancer 2005;103:1631– 43. [21] Uchikado Y, Natsugoe S, Okumura H, et al. Slug expression in the E-cadherin preserved tumors is related to prognosis in patients with esophageal squamous cell carcinoma. Clin Cancer Res 2005;11: 1174 – 80.
Original article
Snail expression is an independent predictor of tumor recurrence in superficial bladder cancers夡 Franck Bruyere, M.D.a,1, Benjamin Namdarian, M.B.B.S.b,1, Niall M. Corcoran, Ph.D., M.D.b, John Pedersen, M.D.c, Jeremy Ockrim, M.D.d, Bryan B. Voelzke, M.D.e, Uttam Mete, M.D.f, Anthony J. Costello, M.D.b, Christopher M. Hovens, Ph.D.b,* a Department of Urology, CHU Bretonneau, Tours, France Departments of Urology and Surgery, Royal Melbourne Hospital, University of Melbourne, Parkville, Australia c TissuPath Pty Ltd., Hawthorn, Australia d Departments of Urology, Imperial College School of Medicine, Hammersmith Hospitals NHS Trust, London, United Kingdom e University of California, San Francisco, CA 94143, USA f Postgraduate Institute of Medical Education and Research, Chandigarh, India b
Received 21 August 2008; received in revised form 6 November 2008; accepted 9 November 2008
Abstract Background: Epithelial-mesenchymal transition (EMT) is known to play an important role in the development of tumor invasion and progression in tumors of epithelial origin. Objectives: Our aim was to investigate the role of Snail transcription repressor family members in human bladder pathogenesis. Material and methods: We evaluated levels of Snail and Slug in 87 patients who received transurethral resection of a transitional cell carcinoma at our institution during the period from June 1999 until November 2003. Immunohistochemistry was performed on tissue microarrays, and expression correlated with pathological variables and clinical outcomes. Degree and intensity of Snail and Slug staining was quantified by immunohistochemistry. Results: There was no apparent enrichment in strong vs. weak staining for either Snail (43.7% vs. 56.3%) or Slug (46% vs. 54%) in the superficial bladder tumors. Univariate analysis determined that tumor focality and Snail expression were significantly associated with tumor recurrence (P ⬍ 0.05). Only for tumor focality did such a relationship exist when assessing tumor progression. Multivariate analysis using the Cox’s proportional hazards model revealed similar results to that of the univariate analysis. Snail expression (P ⫽ 0.038) and tumor focality (P ⫽ 0.011) were independent and significant prognostic factors for tumor recurrence in all patients. However, only tumor focality was an independent predictor of tumor progression (P ⫽ 0.034). Conclusions: High expression of Snail in superficial bladder tumors is a strong predictor of tumor recurrence enhancing risk stratification and prognostication. © 2010 Elsevier Inc. All rights reserved. Keywords: Snail; Slug; Immunohistochemistry; Recurrence; Carcinoma; Transitional cell
1. Introduction
夡 This work was supported by funds from the Melbourne Urology Trust, by a generous gift from Roger Riordan from the Cybec Trust. N.M.C. was supported by a Benjamin Rank Surgical Fellowship from the Royal Melbourne Hospital, and a NHMRC Medical Postgraduate Scholarship. B.N. was supported by the 2007 Raelene Boyle Scholarship from the Royal Australasian College of Surgeons. * Corresponding author. Tel.: ⫹613-9342-7703/4; fax: ⫹613-9347-6488. E-mail address: [email protected] (C.M. Hovens). 1 These authors contributed equally to this work.
1078-1439/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2008.11.005
Cases of bladder cancer contribute significantly to the overall tumor burden with more than 60,000 new cases diagnosed in the United States in 2006, with approximately 13,000 deaths from the disease [1]. The most common bladder cancer histological subtype is that of transitional cell carcinoma, with some 70% to 80% of these cases initially presenting with superficial tumors that are confined to the epithelium or subepithelial connective tissue. Transurethral surgical resection of these tumors remains the treat-
592
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
ment of choice, however 31% up to 78% of these cases will recur [2]. Of those that recur, up to 45% progress to invade the underlying detrusor muscle, which is associated with a significantly poorer outcome [3,4]. Current pathological classification schema cannot adequately predict tumor behavior nor permit patient selection for early aggressive treatment or long-term surveillance as appropriate. Molecular profiling of key tumor signaling and regulatory proteins in bladder cancer may lead to improved tumor classification and clinical outcomes. Transition from an epithelial to mesenchymal state (EMT) is a necessary event prior to tumor cells of epithelial origin migrating and metastasizing to secondary tissues [5–7]. Loss of E-cadherin mediated cell adhesion is a common event in epithelial cancers and has been correlated with dedifferentiation, a higher risk of local invasion and metastatic disease, and poor prognosis in a variety of tumors of epithelial origin [8 –12]. Recent attention has focused on transcription factors of the Snail superfamily, Snail [13] and SLUG [14], which can suppress transcription from the E-cadherin promoter. Expression of the Snail superfamily transcription factors has been found to negatively correlate with E-cadherin expression in various epithelial cancers [15–18]. The aims of this study were to analyze the expression levels of the Slug and Snail transcription repressor proteins family members in superficial bladder tumor specimens and to correlate these levels with clinical and pathological features and oncological outcomes.
2. Materials and methods 2.1. Protocol In our study we assessed 87 patients (70 men and 17 women) with a mean age of 71 years during the period from June 1999 until November 2003, who received transurethral resection of a transitional cell carcinoma at our institution. Approval for the study was granted by the Research Directorate of Melbourne Health, Australia. Samples of the tumors that had been fixed in 10% formalin and paraffin embedded shortly after the procedure were obtained from the archives of the Department of Pathology. Sections of the specimens were stained with hematoxylin-eosin (HE) and analyzed by two independent practitioners, classifying each according to TNM and WHO criteria. 2.2. Tissue microarrays (TMA) generation From the HE-stained sections tumor regions were identified and 2 to 3 tissue cylinders of diameter 1.5 mm were taken and layered into a recipient block using a tissue chip microarrayer. The recipient block was subsequently sec-
tioned onto Histobond (Leica, Marienfeld, Germany) sialinized glass slides. Graded ethanol solutions were used to deparaffinize and rehydrate the 5 m paraffin-embedded specimen sections while 0.3% hydrogen peroxide was used to block endogenous peroxidase activity. Antigen retrieval was performed by heating the slide in a microwave oven in EDTA buffer (10 mM; pH 8, 7 minutes). Sections were cooled at room temperature for 30 minutes before incubated with primary antibodies at 4oC. The following antibodies were used: SNAI 1 (H-130):sc-28199 (1:400); Santa Cruz Biotechnology (Santa Cruz, CA), SLUG (D-19):sc-10437 (1:100); Santa Cruz Biotechnology. Following primary antibody application, the sections were incubated for 30 minutes with peroxidase labeled polymer (DAKO Cytomation Envision⫹ dual link system; Carpinteria, CA). Sections were subsequently washed thoroughly in phosphate-buffered saline, and 3,3=-diamino-benzidine (DAKO Cytomation Liquid DAB⫹) was used to develop the peroxidase signal. In assessing the slides, two independent observers blindly examined the staining in TMAs, and a mean score was calculated. The positive reaction was scored into 4 categories for both the intensity of staining and the percentage of positive cells. Grades according to the intensity of the staining included: 0, 1⫹, 2⫹, and 3⫹ while the percentages of positive cells were separated into: 0 (0%), 1 (1% to 33%), 2 (34 to 66%), and 3 (67 to 100%). In order to maintain consistency between observers, when the rare occurrence of a discrepancy arose, a third observer gave the final score. In the final statistical analysis, the staining results were classified into 2 groups depending on the intensity and the percentage of positive cells. Weak staining (intensity 0 or 1, or more but percent category 0 or 1); and strong staining (intensity 2 or 3 and percent category 2 or 3). 2.3. Statistical analysis In performing statistical analyses with SPSS version 9 (Chicago, IL), age, sex, stage, grade, multifocality, tumor weight, and antibody overexpression were the tumor and patient variables analyzed. Chi Square test or Fisher’s exact test were used to demonstrate any intercorrelations between stage, grade, multifocality, tumor focality, and antibody over expression. Time to recurrence and progression were analyzed using the Kaplan-Meier method. Correlations between tumor recurrence or progression, and antibody overexpression, as well as pathological variables such as grade and focality were identified using the log rank test. The date of the last cystoscopy was used to define the periods of assessment for patients without recurrence, progression, or cancer death. The Cox proportional hazards model was used for multivariate analysis of parameters predictive of tumor recurrence and to estimate relative risks. P values reported in the article are all 2-sided, significant at P ⬍ 0.05.
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
593
Table 1 Clinical and histological data of 87 patients with superficial bladder cancer
Table 3 Tumor recurrence data for Slug and Snail proteins for 87 superficial bladder tumor specimens
Number Sex (M:F) ratio Mean age (y) SD Recurrence (%) Progression (%) Median follow-up Grade (%) PUNLMP Low grade High grade Uni/multifocal Median Tumor weight (range) g
Variable
87 70/17 71 (30–96) 50 (57.5%), mean 14.8 mo 11 (12.6%), mean 16.3 mo 33.9 mo (0.5–66) 12 (13.8%) 61 (70.1%) 14 (16.1%) 61.8 %/38.2% (11 missing) 3 (1–55)
3. Results 3.1. Clinicopathologic summary
Grade PUNLMP Low High Focality Unifocal Multifocal Weight ⬍Median ⬎Median Slug Poor Strong Snail Poor Strong
Number recurred
Mean time to recurrence (SE)
9/12 33/61 8/14
28.3 (16.7–39.8) 29.4 (23.3–35.6) 29.5 (15.9–43.1)
23/47 20/29
33.6 (27.0–40.2) 19.9 (11.5–28.3)
28/46 22/41
29.4 (22.6–36.2) 27.7 (21.0–34.4)
26/47 24/40
27.7 (21.4–34.1) 29.5 (22.2–36.8)
24/49 26/38
33.6 (27.0–40.2) 22.6 (15.7–29.6)
Lag rank
P value
0.04
0.9795
7.34
0.0068
0
0.9457
0
0.994
4.85
0.0275
Cox proportional hazards model
A total of 87 consecutive cases of superficial urothelial transitional cell carcinoma of the bladder were selected in this study (Table 1). All tumors were classified as superficial pTa tumors according to the TNM criteria, and subgrouped into grades by the World Health Organization (WHO) criteria with papillary urothelial neoplasms of low malignant potential (PUNLMP) comprising 12 patients, low grade consisting of 61 patients, and high grade consisting of 14 patients (Table 1). Interestingly, we were unable to find a correlation on univariate or multivariate analyses between tumor grade (Table 2) and recurrence (Table 3) or progression (Table 4), which we believe reflects disease presentation in our cohort, the current WHO grading system placing a larger proportion of cases into the low grade category, hence limiting the power of any statistical comparison to PUNLMP or high grade tumors. Further clinical and histologic data are summarized in Table 1. 3.2. Tissue immunohistochemistry analysis We examined the expression of both Snail and Slug in the 87 tumor samples (Fig. 1). There was no apparent enrichment in strong vs. weak staining for either Snail (43.7% vs. 56.3%) or Slug (46% vs. 54%) in the superficial bladder tumors (Fig. 2). We then correlated protein expres-
Table 2 Pearson 2 odds ratio correlation between Slug and Snail expression and standard bladder tumor prognostic indicators
Slug Snail
Weight
Focality
Grade
1.2 (0.5–2.9) P ⫽ 0.62 1.48 (0.63–3.47) P ⫽ 0.365
2.3 (0.9–6.1) P ⫽ 0.087 0.91 (0.36–2.32) P ⫽ 0.846
NA P ⫽ 0.936 NA P ⫽ 0.54
Variable
HR (95% CI)
P value
Grade Weight Focality Slug Snail
1.06 0.93 2.25 1.07 1.87
0.83 0.81 0.011 0.83 0.038
(0.65–1.73) (0.53–1.65) (1.20–4.21) (0.58–1.98) (1.04–3.38)
sion with established histological predictors of bladder tumor aggression, including grade, tumor bulk, and focality. We established on Pearson 2 odds ratio correlation that neither Snail nor Slug expression levels correlated significantly with any of these histologic predictive parameters (Table 2). We then proceeded to examine on univariate and multivariate analysis whether protein expression of Snail or Slug correlated with patient prognosis in regards to tumor recurrence and progression in all patients (n ⫽ 87). The univariate analysis using the log rank test determined that tumor focality (P ⫽ 0.0068) and Snail expression (P ⫽ 0.0275) were the only parameters significantly associated with tumor recurrence (Table 3). The predictive ability of Snail is demonstrated in Fig. 3, showing the Kaplan Meier survival curve. Multivariate analysis using Cox’s proportional hazards model revealed that only Snail expression (P ⫽ 0.038) and tumor focality (P ⫽ 0.011) were independent and significant prognostic factors for tumor recurrence in all patients (Table 3). We then determined if expression of these transcriptional repressor proteins, influenced tumor progression. We found that tumor focality (P ⫽ 0.0237) but neither Snail nor Slug levels significantly predicted for tumor progression on univariate analysis (Table 4). Upon multivariate analysis, again only tumor focality was an independent predictor of tumor progression (P ⫽ 0.034, Table 4).
594
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
Table 4 Tumor progression data for Slug and Snail proteins for 87 superficial bladder tumor specimens Variable Grade PUNLMP Low High Focality Unifocal Multifocal Weight ⬍Median ⬎Median Slug Poor Strong Snail Poor Strong
Number progressed
Mean time to progression (SE)
3/12 5/61 3/14
49.1 (38.5–59.6) 60.9 (56.7–65.2) 50.4 (38.9–61.9)
3/47 7/29
58.5 (54.6–62.3) 51.9 (42.2–60.6)
5/46 6/41
60.1 (55.2–64.9) 49.1 (43.4–54.9)
6/47 5/40
58 (52.1–63.9) 53.8 (48.8–58.9)
5/49 6/38
53.2 (49.2–57.1) 56.3 (49.2–63.4)
Log rank
P value
2.9
0.234
5.12
0.0237
0.71
0.4
0.04
0.843
0.71
0.399
Cox proportional hazards model HR (95% CI)
P value
Grade Weight Focality Slug Snail
0.75 1.64 4.56 1.08 1.55
0.63 0.5 0.034 0.9 0.5
(0.23–2.43) (0.49–5.47) (1.12–18.57) (0.31–3.82) (0.44–5.29)
4. Discussion Currently, decisions regarding treatment regimens for urothelial carcinoma are based solely on the TNM staging schema. This classification criteria is well suited to describe tumor morphology, however, this information cannot be extrapolated to provide a full assessment of the risk for individual tumor recurrence, progression, or response to a given course of treatment. Panels of molecular biomarkers, which have the potential to be objectively quantified and routinely incorporated into pathological analyses, might overcome current morphological assessments and permit eventual stratification of clinically relevant risk profiles of pathologically similar bladder tumors. We determined the expression profiles of Snail and Slug and their relationship to the clinicopathological features of the superficial bladder tumors. We found that elevated levels of Snail expression were significantly associated with bladder tumor recurrence, which is suggestive that this marker may be an indicator of poor prognosis in superficial bladder cancer. Surprisingly, we found no such correlation with the Snail related transcriptional repressor, Slug. The finding that Snail expression is associated with tumor recurrence conforms to the hypothesis that transcriptional regulators of EMT are important for advanced tumor features such as invasion and metastasis. Interestingly, in their normal physiologic context, Snail genes appear to control cell movement processes that do not require a full EMT
Fig. 1. Representative images of immunohistochemical staining of the indicated antibodies on superficial bladder cancer microarrays. Magnification at 100⫻ to 400⫻. (Color version of figure is available online.)
such as mesoderm formation by regulating cell adhesion and migration [19]. Snail expression is positively correlated with poor outcome in metastatic breast cancer patients, with Snail RT-PCR positive breast carcinoma effusions associAntibody Slug Snail
Strong Staining 46 43.7
Weak Staining 54 56.3
100% 90% 80% 70% Percentage Staining
Variable
60% Weak
50%
Strong
40% 30% 20% 10% 0% Slug
Snail
Fig. 2. Expression of Snail and Slug antibodies as assessed by immunohistochemical staining of the superficial bladder tumor cohort. (Color version of figure is available online.)
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
595
[16]. Overall, these findings suggest that different members of the Snail transcription repressor family are likely to have highly specific roles in regulating EMT in different tumor contexts. A knowledge of these roles will need to be obtained before specific therapies, targeted to individual members of this transcriptional repressor family, can be effectively applied to treat different epithelial carcinomas.
5. Conclusions To date, no single biomarker has consistently enabled risk stratification and prognostication of superficial bladder tumors. When analyzing Snail and Slug expression levels, we have found that high expression of Snail in superficial bladder tumors significantly predicts tumor recurrence in these patients.
References
Fig. 3. Kaplan-Meyer curves of tumor recurrence analyzing Slug and Snail expression levels in superficial bladder tumors.
ated with extremely poor survival [20]. In addition, a recent analysis of 48 primary ovarian tumors and their corresponding metastases for expression of E-cadherin and Snail by immunohistochemistry found that Snail expression in metastases (P ⫽ 0.047) but not the primary tumor was significantly associated with reduced overall survival of the patients [17]. However, neither Snail nor Slug expression were positively correlated with metastatic effusions, disease-free survival, or overall survival in an ovarian cancer cohort [20]. In this tumor setting, it was the Snail transcription factor family member, Smad interacting protein one (sip1), that was associated with higher tumor stage, metastatic effusions and poorer overall survival [20]. In contrast to our findings with superficial bladder cancers, Slug up-regulation has been found to correlate with deeper invasion, lymph node metastases and tumor stage in esophageal squamous cell carcinoma [21], and with distant metastases and advanced tumor stage in gastric carcinoma
[1] Jemal A, Siegel R, Ward E, et al. Cancer statistics 2006. CA Cancer J Clin 2006;56:106 –30. [2] Babjuk M, Oosterlinck W, Sylvester R, et al. EAU Guidelines on Non-muscle-invasive urothelial carcinoma of the bladder. Eur Urol 2008;54:303–14. [3] Holmang S, Hedelin H, Anderstrom C, et al. The relationship among multiple recurrences, progression and prognosis of patients with stages Ta and T1 transitional cell cancer of the bladder followed for at least 20 years. J Urol 1995;153:1823– 6, Discussion 1826 –7. [4] Heney NM, Ahmed S, Flanagan MJ, et al. Superficial bladder cancer: Progression and recurrence. J Urol 1983;130:1083– 6. [5] Bagnato A, Rosano L. Epithelial-mesenchymal transition in ovarian cancer progression: A crucial role for the endothelin axis. Cells Tissues Organs 2007;185:85–94. [6] Moustakas A, Heldin CH. Signaling networks guiding epithelialmesenchymal transitions during embryogenesis and cancer progression. Cancer Sci 2007;98:1512–20. [7] Hugo H, Ackland ML, Blick T, et al. Epithelial-mesenchymal and mesenchymal-epithelial transitions in carcinoma progression. J Cell Physiol 2007;213:374 – 83. [8] Moll R, Mitze M, Frixen UH, et al. Differential loss of E-cadherin expression in infiltrating ductal and lobular breast carcinomas. Am J Pathol 1993;143:1731– 42. [9] Fujimoto J, Ichigo S, Hirose R, et al. Expression of E-cadherin and ␣and -catenin mRNAs in uterine cervical cancers. Tumor Biol 1997; 18:206 –12. [10] Jawhari A, Jordan S, Poole S, et al. Abnormal immunoreactivity of the E-cadherin-catenin complex in gastric carcinoma: Relationship with patient survival. Gastroenterology 1997;112:46 –54. [11] Jeffers MD, Paxton J, Bolger B, et al. E-cadherin and integrin cell adhesion molecule expression in invasive and in situ carcinoma of the cervix. Gynecol Oncol 1997;64:481– 6. [12] Huiping C, Kristjansdottir S, Jonasson JG, et al. Alterations of E-cadherin and -catenin in gastric cancer. BMC Cancer 2001;1:16. [13] Cano A, Perez-Moreno MA, Rodrigo I, et al. The transcription factor Snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression. Nat Cell Biol 2000;2:76 – 83. [14] Nieto MA. The Snail superfamily of zinc-finger transcription factors. Nat Rev Mol Cell Biol 2002;3:155– 66. [15] Rosivatz E, Becker I, Specht K, et al. Differential expression of the epithelial-mesenchymal transition regulators Snail, SIP1, and twist in gastric cancer. Am J Pathol 2002;161:1881–91.
596
F. Bruyere et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 591–596
[16] Castro Alves C, Rosivatz E, Schott C, et al. Slug is overexpressed in gastric carcinomas and may act synergistically with SIP1 and Snail in the down-regulation of E-cadherin. J Pathol 2007;211: 507–15. [17] Blechschmidt K, Sassen S, Schmalfeldt B, et al. The E-cadherin repressor Snail is associated with lower overall survival of ovarian cancer patients. Br J Cancer 2008;98:489 –95. [18] Becker KF, Rosivatz E, Blechschmidt K, et al. Analysis of the E-cadherin repressor Snail in primary human cancers. Cells Tissues Organs 2007;185:204 –12.
[19] Barrallo-Gimeno A, Nieto MA. The Snail genes as inducers of cell movement and survival: Implications in development and cancer. Development 2005;132:3151– 61. [20] Elloul S, Elstrand MB, Nesland JM, et al. Snail, Slug, and Smadinteracting protein 1 as novel parameters of disease aggressiveness in metastatic ovarian and breast carcinoma. Cancer 2005;103:1631– 43. [21] Uchikado Y, Natsugoe S, Okumura H, et al. Slug expression in the E-cadherin preserved tumors is related to prognosis in patients with esophageal squamous cell carcinoma. Clin Cancer Res 2005;11: 1174 – 80.